CODING SCHEME FOR THE ANALYSIS OF CLASSROOM DISCOURSE

Gordon Wells with DICEP

This document has been compiled over a considerable period of time as the DICEP group has moved towards a specific focus for the analysis of the videorecorded and transcribed observations that have been made in participating teachers' classrooms. Part 1 contains a proposal for the general orientation to be adopted in investigating classroom discourse. Part 2 outlines the basic structure of the proposed scheme of analysis. Part 3 contains the specific coding framework developed for the investigation. PART 1. NOTES ON THE RELATIONSHIP BETWEEN VISION, THEORY AND PRACTICE 1. The stated goals of our project make it clear that our interest in classroom discourse is neither purely theoretical, nor an attempt to produce a descriptive survey of what happens in classrooms. As is implied by the term 'action research' to describe our collaborative activities, our goals are improvement in action and in understanding, achieved through a dialectic between them. This means that our project is essentially about change: about reflecting on what is currently happening, imagining possibilities for improvement, acting to achieve them, evaluating the outcome, and increasing understanding by trying to make sense of the whole cycle in terms of some overarching theory(s). Then continuing with a further, related, cycle. 2. This involves us in working on a number of levels - or facets - simultaneously. We refer to these as: 'Vision', 'Practice', 'Theory', 'Data'. Each level illumines, and is illumined by the others; and, at each level, there is a continuing interaction between using the evidence we are collecting for knowledge building, and putting the understanding achieved in this process to good effect in guiding action. 3. Vision. This is what, for each of us, energizes and guides our practice. It can be thought of in terms of a number of goals, which include: a) creating communities characterized by: inclusiveness, equity, caring, as well as by intellectual achievement;

b) giving a high priority to knowledge building and understanding through inquiry, while not neglecting the routine processes and skills needed to engage in them; c) encouraging collaboration - between teacher and students, as well as among students; valuing and building, whenever possible, on students' contributions to the activity in progress, so that knowledge is coconstructed, rather than delivered; d) broadening interests and recognizing and valuing the contributions of 'experts' beyond the classroom; bringing the classroom community into a two-way relationship with communities (local/world-wide, practical/intellectual) beyond the classroom by participating in their practices; e) acknowledging and taking into account that, whatever the activity, the whole person is always involved (body/mind/feelings/dispositions); f) providing for individual growth and self-determination as well as for the classroom community as a whole; 4. Practice. Although we can work towards agreement about the vision, we necessarily differ with respect to practice because of the unique combination of conditions and participants that we each work with. But there is another crucial difference between vision and practice. Vision is essentially abstract and ‘synoptic’; practice is ‘dynamic’, realized in the successive actions and decisions of particular individuals and communities in relation to the possibilities and constraints of their material, social and intellectual environments. Although there is clearly a relationship between vision and practice, it is not a simple one, such as is suggested by the term ‘implementation’. However, it is towards understanding this relationship, as well as towards optimizing practice, that our individual and collaborative inquiries are directed. 5. Theory. Theory has a dual function. On the one hand, it is what is produced in, and is the outcome of, attempting to make sense of one's own and other people's experience in a coherent and systematic manner. In this function, it is a distillation of the diversity of particular, dynamic experiences in a synoptic model that attempts to explain by reference to more general categories and relationships. A theory is an ‘artifact’ created in the process of knowledge building in collaboration with others. On the other hand, a theory is only valuable to the extent that it enables us to better understand the situations we find ourselves in so that we can act more effectively. In this sense, a theory is a tool for use in understanding and action. 6. Data are also artifacts; they are specific outcomes of particular activities that, deliberately collected, represent some aspect(s) of those activities. The important thing is that data continue to exist after the activity is over and can therefore be revisited and analysed. Our data consist of videorecordings of classroom activities, texts of various kinds, interviews, email discussions, and recordings of parts of out meetings. 7. Data stand to theory rather as practice stands to vision. Where vision and theory are abstract and synoptic, practice and data are specific and dynamic. However, there is another way of considering them: theory and data, considered as a functional sytem, can be thought of as a tool that mediates understanding of the relationship between vision and practice.

If this is correct, the theory we use should be compatible with our vision; the explanations it offers of particular instances of practice should be interpretable in terms of the vision. By the same token, the data we focus on should represent central aspects of practice and, when analyzed in terms of the theory, should allow us to judge how far practice is truly realizing the vision. 8. One of the strengths of our overall model is that it is not unidirectional. Investigations of practice can lead to clarifications and enrichments of vision, and analyses of data can lead to development and/or modifications of the theory, as well as vice versa. Equally, theory and vision influence each other, just as, on the one hand, analyses of data help us to see what is actually happening in practice, and on the other, focusing on particular aspects of practice calls for the development of new techniques of analysis derived from the theory. Assumptions Underlying the Coding Scheme In devising a scheme for the systematic analysis of discourse, it is important to make explicit the assumptions on which it is based. The following are some issues for discussion. 1. The scope of ‘discourse’ The term ‘discourse’ is used in different ways by different people and on different occasions. In the present context, ‘discourse’ refers to communication that is both dialogic and linguistically based. It therefore includes written discourse as well as spoken, and inner as well as social discourse. However, in both phylogenetic and ontogenetic development, face-to-face interaction precedes discourse in other modes, and speech serves as the medium in which other modes of meaning-making are most frequently planned, interpreted and discussed. For this reason, the following assumptions are formulated with respect to spoken discourse. They will certainly need to be modified when considering other modes. 2. Episodes of discourse occur as means to the achievement of activity goals In general, discourse in the classroom, as in other settings, does not occur as an end in itself, but as a means of carrying out some activity(s) in which the participants are jointly involved. It is best understood, therefore, as functioning within a larger framework of mediated social activity (Wertsch, 1994). On this basis, it is proposed (Wells, 1993, 1996) that the scheme of analysis should be constructed on the basis of an articulation of activity theory (Leont'ev, 1981; Engestrom, 1991) and systemic linguistics (Halliday, 1978, 1993). Furthermore, spoken discourse is only one of several mediational means, including action with material objects, drawing, reading, writing, etc., that are employed to achieve the goals of classroom activities. Episodes of spoken discourse can therefore be thought of as co-occurring or alternating with, and thereby complementing, other forms of semiotic behaviour in operationalizing the tasks that make up those activities. The relationship between discourse and the goal of activity varies. In some cases, the activity

goal is constituted and achieved solely through the discourse (e.g. constructing a theoretical explanation of something); in other cases, a proposed, ongoing, or past activity forms the topic of the discourse; in yet others, the discourse takes place in parallel with a non-verbal activity, which is the primary focus of attention. In all cases, however, discourse may also play a ‘meta’ role in (re)negotiating the goal of the activity or task, and in monitoring and evaluating progress toward it. 3. Discourse is necessarily collaborative; but this does not imply agreement Discourse is necessarily collaborative - at least to some degree. Individual contributions are only effective if they mesh with co-participants’ interpretations of the situation, including the activity in which they are jointly engaged and the preceding discourse contributions. However, this does not mean that there is uniformity of purpose or of interpretation. In other words, while intersubjective agreement is normally the aim of interaction, it is not a prerequisite condition for collaboration. Indeed, without some divergence or disagreement, there would be no increase in individual or shared understanding (Matusov, 1996). In multi-party talk, particularly where there are inequalities of status or power, there may be multiple purposes, some of which may be in conflict. Furthermore, as would be expected from a constructivist perspective, participants’ interpretations of others’ meanings are always coloured by their own current preoccupations as well as by their personal construal and knowledge of the topic, and of what is said and done, based on their past experience. Even discourse with self is collaborative, when the larger context of activity to which it relates is taken into account. This is true not only for the means used - the social language from which the medium of inner speech is derived. It also holds for the functions of inner discourse, which are often both dialogic in nature and oriented towards previous or subsequent direct participation in social activity. 4. Discourse is both emergent and generically structured Because discourse is progressively constructed by multiple participants, move by move, its meaning and structure are constantly emerging. Each move takes account of what preceded and, in turn, sets up affordances and constraints for what will follow. At the same time, no move can fully control what follows; its import for the discourse is only seen in the way in which it is taken up in subsequent moves. However, this does not mean that there is no larger organizational structure. Drawing on Bakhtin’s (1986) work and on current genre theory (Kamberelis, 1995), discourse can be thought of as being made up of segments on different scales, from minimal exchanges to lengthy episodes, each based on a generic organizational pattern of constituent parts that is the culturally accepted way of carrying out the task or activity which it serves to operationalize. To a considerable extent, making an effective move is dependent on recognizing which genre or micro-genre is currently being co-constructed and making a contribution which is interpreted as being appropriate in this context. The degree of constraint on what move can be made at successive points, and hence the extent to which the larger structure is predictable, depends on the nature of the activity, the extent to

which there is a specific genre associated with it, and the degree to which one participant exercises control over the way in which the discourse unfolds. Teacher-directed, whole-class ‘recitation’ discourse, for example, tends to be organized in a hierarchical-sequential manner according to the principle that one unit is completed before the next one starts, with the teacher both initiating and completing each unit and determining who may participate and the nature of the contributions that will be accepted. However, in less controlled situations, and where there is greater equality among participants, the sequential organization is constantly under negotiation and is sometimes contested. Furthermore, more than one person may speak at the same time and there may be more than one conversation going on simultaneously within the same group. There are also occasions when one unit may be embedded within another (e.g. to negotiate problems in the interaction). 4. Meaning is situated in the ongoing activity, which is linked to other related activities within the culture and within the experience of each of the participants The meanings made in discourse are always situated with respect to the particular activity in which they occur and to the point reached in that activity. Discourse moves thus cannot be taken as context-free expressions of the speaker’s ‘true’ beliefs and attitudes, but must be understood as strategic contributions, designed to advance the current discourse towards the goal envisaged by the speaker (which may not be the same as that envisaged by other participants). Meanings are also strongly influenced by the connections made by participants to related experiences, both personal and collective. These exist on several time-scales: within the current activity/discourse; within the participants' individual and collective experience of similar or related activities in their community; within the history of the activity in the culture more generally. Some of these connections concern the ‘content’ of the activity; others relate to the ‘ground-rules’ for participation (Edwards and Mercer, 1987). Connections are also made to other texts, both spoken and written - to the information they contain, and to the responses they evoke. These various types of connection and the constitutive role they play in the ongoing process of collaborative meaning making are referred to in terms of ‘intertextual relations’ (Lemke, 1994). 5. Discourse proceeds simultaneously on multiple dimensions of meaning Participation in discourse involves the whole person, body and heart as well as mind. Moves not only refer to aspects of the world and theories about the world, they also express the speaker’s involvement in and attitude to this ‘content’, as well as adopting an orientation to the other participants with respect to both content and involvement. (To complicate matters, the same move may be intended to be taken up differentially by different co-participants.) Moves are also strategically shaped to fit the generically defined slot in which they are placed in the unfolding discourse (i.e. as answer to a question, or as ‘another anecdote of the same kind’). Only some of this complex web of meaning is realized lexico-grammatically (although this tends to be the only semiotic modality that is adequately captured in a written transcription). Other significant modalities include various parameters of intonation, pacing and pausing, gaze, gesture and spatial orientation to the people and objects involved in the activity.

When considering the functions of moves, therefore, it is not adequate to describe only the function encoded in the lexico-grammar, and still less to focus only on the ‘ideational’ function. Implications The implications of basing analysis on these assumptions seem to include the following: 1. We should try to ensure that written representations of actual discourse data are as ‘thick’ as we can make them by: a) providing as much detail as possible about the participants, the activity and the historical, social and intellectual context; b) including information about the modalities of meaning-making that are not represented lexico-grammatically. 2. While recognizing and attempting to convey the uniqueness of each event, we should also attempt to describe events in terms of a common framework and make use of a common set of analytic categories. 3. We should accept that there is no single, correct, analysis of any event. Like moves in discourse, analytic codings or descriptions are always situated and depend on the analyst’s purpose and perspective. In addition, multi-party events can be viewed from the perspective of more than one of the participants and thus give rise to alternative, but complementary analyses. Since it is impossible to do justice to all perspectives, it is important to make clear what perspective is adopted and to recognize the limitations and/or bias that this entails.

PART 2. UNITS OF ANALYSIS AND THE RELATIONSHIP BETWEEN THEM Table 1 shows the relationship of discourse to activity. In educational settings, it is convenient to use the more familiar terms of ‘Curricular Unit’, ‘Activity’, and ‘Task’ to label the units of activity in the classroom. These correspond to the level of ‘action’ in activity theory. Again to avoid confusion, we use the term ‘Practice of Education’ to refer to the highest level of activity theory. Each teacher has her/his own vision of what the practice of education should be chiefly concerned with and it is to realize this vision that s/he selects the topics and approaches used at the level of curricular unit, activity and task. But it is ultimately through the ‘operations’ of meaning making in which the community and its individual members engage that the vision is put into practice. It goes without saying that this section presents a highly synoptic representation. In reality, the relationship is as much bottom-up as top-down, and it is always to a considerable degree emergent. Table 1. The Enactment of the Practice of Education ‘ACTIVITY’ ‘ACTION’ ‘OPERATION’ (Motive) (Goal) (Means) Practice of (a) cultural Curricular Increasing mastery of: Education reproduction Unit (a) content knowledge (b) development of (b) discipline-based individual practices potential (c) tools & artifacts (c) fostering of (d) metacognition communities of (e) collaboration inquiry etc. Curricular Outcomes related Use of semiotic tools, Activity to (a)-(e) above including spoken discourse , e.g. Curriculum genres (cf. Christie, 1993) Task Completion of a e.g. Co-construction component of an of episode of discourse activity outcome Step Contribution to E.g. Co-construction outcome of task of sequence of discourse, using a microgenre, e.g. triadic dialogue

The Constituency Relationship Between Units of Spoken Discourse Table 2, and the accompanying figure 1, show the hierarchical relationship between units of discourse. Each unit consists of at least one unit of the level below. In practice, however, higher level units are always built up over real time from units at levels below. Table 2. The Constituency Relationship Between Units of Spoken Discourse Episode: Typically corresponds in scope to a recognizable task within a curricular activity. Sequence: All the moves required to fulfill the expectations set up by the initiating move in the nuclear exchange around which it is organized. Exchange: The minimal unit of interaction, it consists of an initiating move and a responding move; in certain types of discourse, there may also be a follow-up move. Move: A contribution to an exchange made by a participant in a single speaking turn. (Note: A turn may involve moves that contribute to more than one exchange, and a move may also function as an (implicit)response in one exchange and as the initiation of the following exchange.) Act: A move, or one of its constituents, viewed from the perspective of the speech function it realizes.

Figure 1. The Hierarchical Relationship Between Units of Discourse

Episode

Sequence 1 ... (Sequence n)

(Preparatory Exch) Nuclear Exch (Dependent Exch) (Embedded Exch)

Initiating Move Responding Move (Follow-up Move)

Segmentation into Units In order to carry out systematic analysis, even of a simple kind, it is necessary to segment the stream of speech (or its transcribed representation) into functional units - at least roughly. Participants must do this (in an informal and barely conscious manner) in order to know where they have got to in the conversation. But they do it in the moment, with only their own interpretation of what has already been said to base their judgments on. The analyst is in the privileged position of being able to view the whole conversation, as it were synoptically from above, and to see how particular moves were subsequently taken up (or not) in subsequent moves. While for some purposes it is not necessary to code every episode in detail, it is probably useful to mark boundaries between sequences and clusters of related sequences in order to get a general idea about how the episode developed. However, when coding in detail it is a good idea to adopt the participants' perspective as the primary one when deciding where a unit ends and a new one begins; of course the analyst also has the ability to look ahead to see whether what was expected actually transpired. Nevertheless, in practice - despite the categorical nature of the descriptions below - boundaries between units are not always clear-cut. Nor is it always easy to decide whether a particular move is part of an exchange/sequence already in progress or the beginning of a new one. A good test is to ask how the participants treated the decision. In the following notes, units are arranged in descending order of size and scope. In conversation, participants construct one move at a time on the basis of what has gone before. In certain respects, coding works from the larger units down. However, it is important to remember that coding is an attempt to represent what the participants were understanding and doing. An episode consists of the interaction that takes place in carrying out some recognizable task or sub-task. In general, episode boundaries are signalled by a change of task (often announced, in whole class activities) or by a change of participant structure. For this reason, episodes often begin and end with boundary exchanges (which may, of course, consist of a single initiating move, to which no-one bothers to respond, or at least not verbally). Episodes consist of an indefinite number of sequences which are linked either by continuity of topic or of the action to which they refer. In some cases, the episode may consist of a number of clusters of sequences, each of which has internal coherence, and is related to other clusters, either by the sequential organization of the practical task or the larger discourse task in which the participants are engaged. This is where the concept of 'action type' or 'curriculum genre' can be helpful. A sequence is organized around its nuclear exchange. In the initiating move of an exchange, a speaker proposes an interactional purpose with respect to some form of information or goods or service. In the simplest case, this purpose is completed in the responding move, or in the follow-up move when one occurs. However, there may be additional exchanges that facilitate or contribute to the achievement of this purpose: either by preparing for it; elaborating or qualifying it; or by dealing with various types of difficulty - of hearing, reference, identification, etc. A sequence may thus be defined as a nuclear exchange and all exchanges that are bound to it (i.e. exchanges that cannot stand on their own but take on their meaning and function in relation to the

nuclear exchange). Sequence boundaries also tend to be marked by the fact that, typically, a sequence starts with a move uttered with intonation relatively high in the speaker’s pitch range; then, over the sequence as a whole, subsequent moves will tend to be lower in pitch, with the final move lower in pitch than those that have preceded. An exchange consists minimally of an initiation move and a response move and, in certain types of exchange, of a follow-up move; these moves are organized in terms of 'commodity' and 'prospectiveness'. The commodity is what is 'exchanged', either 'information' or 'goods and services'; and prospectiveness concerns the extent to which the current move sets up expectations for what follows or meets expectations already set up (Halliday, 1984). There are three basic levels of prospectiveness - high, mid, low - corresponding, respectively, to the speech roles of demanding, giving and acknowledging. These generalized roles map on to the three move types of Initiate, Respond and Follow-up. Where the initiating move makes a demand, a give is expected as a response and this, in turn, is usually followed by a follow-up move that, minimally, acknowledges what has been given. Where the initiating move takes the form of a give, the response move expected is an acknowledge and there is unlikely to be a follow-up move. Exchanges can be of four types. Every sequence contains a nuclear exchange. Sometimes, as when children bid and/or the teacher nominates, the nuclear exchange is preceded by a preparatory exchange. Dependent exchanges may extend or modify some part of the nuclear exchange, by giving details of when, where or why; they may also be used to request a justification or clarification of the response move. Embedded exchanges interrupt an ongoing exchange, usually to deal with problems of communication. Moves consist of one or more acts. Typically, one of these acts will realize a combination of the generalized commodity and speech roles referred to above in a way that is appropriate to the sequential position of the move in the exchange, e.g. 'question' (demand-information) and 'answer' (give-information-on-demand), 'statement' (give-information-without-demand), 'request' (demand-goods/services), 'offer' (give-goods/ services-without-demand), etc. Additional acts may amplify, qualify or justify the 'head' act in some way. Within a move, acts may be realized either consecutively or simultaneously. Acts concerning the commodities of information or goods/services are typically realized consecutively, e.g. a request followed by a justification. Acts involving attitude, affect, etc. are typically realized simultaneously with the former. N.B.: This is an oversimple account of acts, moves and exchanges, as it focuses almost exclusively on the organization of the 'ideational' function of interaction, i.e. what the talk is ostensibly about. However, other 'commodities' such as status, power, attitude, affect and so on, are also at issue. How these map on to the sequential organization is still very much in need of systematic investigation. Subcategorization At each level, categories are more delicately subcategorized according to the specific purpose of the proposed analysis.

PART 3. CODING FRAMEWORK FOR THE ANALYSIS OF THE USE OF FOLLOW-UP MOVES The following framework has been developed to permit investigation of the contexts in which different options are selected to realise the Follow-Up move in triadic dialogue and of the consequences of these selections for the subsequent discourse. A deliberate aim of several of the contributing teachers has been to promote a more collaborative form of knowledge building in the classroom through the co-construction of "progressive discourse", in which all participants build on each other's contributions towards a common understanding of the topic under consideration. One of the principal questions to be investigated is how the selection of different follow-up options may contribute to the achievement of this aim. The revised version (2001) incorporates categories from Nystrand et al. (2000) that address the cognitive demand of questions and the extent of uptake in the Follow-up move. The coding framework is designed to permit selected episodes of videorecorded and transcribed observations to be coded as a succession of "records", where each record corresponds to an exchange. All exchanges are coded, whether or not they contain a follow-up move; however, where a follow-up move occurs, it is coded in some detail, with provision for up to three different acts. In order to allow occurrences of follow-up moves to be related to what precedes and what follows, each record contains information about the episode and sequence in which the exchange occurs; where a follow-up move occurs, the exhange is also coded for its sequel in the immediately following discourse. The remainder of this section consists of two parts: the coding instrument that is used by the coders, and a more detailed definition of the subcategories, where this is considered necessary. The Coding Instrument The instrument assumes that coding will be carried out by (a) segmenting the transcript into its successive constituent exchanges, and then (b) for each exchange, entering the appropriate code in each of the numbered columns in the coding protocol corresponding to the categories in the coding instrument.

Column Field Categories 1 Episode # 1 - n 2 Episode Task M Math Discussion P Science Practical S Science Discussion D Science Presentation O History Practical H History Discussion R History Role-Play L Literature Discussion B Reading Discussion W Writing Discussion T Show and Tell 3 Participant C Whole Class Structure G Group D Dyad 4 Episode Activity C Commenting Orientation O Organizing P Planning T Reporting S Problem-Solving G Generating L Launching B Constructing F Formulating M Monitoring R Reviewing 5 Sequence # 1 - n 6 Sequence Start # Line number of Sequence start

7 Episode N New Topic Development F Further A Adds E Extends S Consolidates C Challenges R Shifts to related. O Relationship opaque

8 Student Link A Adding C Challenging E Extending O Reference opaque 9 Exchange Type N Nuclear D Dependent M Embedded P Preparatory &N Additional Nuclear 10 Cognitive Demand R Rote recall P Previous Conversation M Memory/Prior Knowledge G Generalization A Analysis/Explanation S Speculation 11 Exchange T Teacher Initiator S Same student as involved in previous exchange/sequence N New student 12 Initiation D Demand Prospectiveness G Give G+ Give Plus 13 Demand D Action Subcategories F Full Clause (or more) FI Full (Implicit) C Single Constituent CI Single Constituent (Implicit) A Choice of Alternatives AI Choice of Alternatives (Implicit) P Polar (Yes/No) PI Polar (Implicit) 14 Give E Extended Response Subcategories C Single Constituent A Alternative P Polar (Yes/No)

15 Initiation Function Information: Assumed Known F Fact A Rule-governed answer J Conventional explanation R Report of public event L Connection Information: Personal E Experience I Imagination N Personal Opinion K Exclamation Information: For Negotiation O Opinion P Prediction X Explanation C Conjecture B Connection Goods & Services: Assumed Known D Act Goods & Services: Personal G Intention Goods & Services: For Negotiation S Suggestion Q Clarification Request 16 Responder T Teacher S Same student as in previous exchange N New Student P Pivot (Responds by Initiating) F Feedback (Responds with Follow-up) Q Clarification Request Z Clar.-> Return to Nuclear Exchange 17 Response Function K Acknowledge Y Confirm/Disconfirm U Stall W Exclamation

Information: Assumed Known F Fact A Rule-governed answer R Report of public event J Conventional explanation L Connection Information: Personal E Experience I Imagination N Personal Opinion

Information: For Negotiation O Opinion P Prediction X Explanation C Conjecture B Connection Goods & Services: Assumed Known D Act Goods & Services: Personal G Intention Goods & Services: For Negotiation S Suggestion Q Clarification Request

Z Give Clarification M Cite text (e.g. cite a question)

18 Response Length 1 Minimal (less than 1 clause) 2 Main clause (+ dept. clause) 3 Three or more clauses 19 Follow-up T Teacher S Same student as in previous exchange N New Student A Other Adult

20 Follow-up A Acknowledge Prospectiveness Y Accept R Reject G Give D Demand G+ Give+ 21 Evaluation H Accept + Uptake Level P Accept + Praise Q Follow-up Question J Reject + Justification L Accept/Reject N Null Evaluation 22 F1 Prosp. A Acknowledge G Give D Demand 23 F1 Commodity V Evaluation J Justification/Explanation C Comment U Clarification A Action M Metatalk 24 F1 Subcat. (Contingent on choice of Commodity) Evaluation A Accept R Reject C Correct F Reformulate D Counter Y Repeat (accept) N Repeat (reject) P Praise Comment E Exemplification A Amplification C Connection S Summarise O (Demand) Opinion D (Demand) Description L (Demand) Observation

Action N Now F Future S Suggestion Clarification R Repetition I Identification C Confirm Metatalk C Metacognitive O Metaorganizational 25 F1 Length 1 Minimal 2 Main Clause (+ dept. clause) 3 Three clauses or more 26-28 F2 (as for F1) 29-32 F3 (as for F1) 33 Sequel Type M Embedded Exchange contin. D Dependent Exchange contin. F Sequence: Further A Sequence: Adds E Sequence: Extends S Sequence: Consolidates C Sequence: Challenges R Sequence: Shifts to related O Sequence: Relationship opaque N New Topic Z End of Episode

Definitions of Selected Categories and Sub-Categories

Only those categories judged to need further explanation are included below. The number at the beginning of each sub-section corresponds to the relevant column in the instrument above. The coding framework assumes the theoretical perspective outlined in section 2 above, in which the majority of events that make up life in classrooms are seen as the goal-directed activities through which the Practice of Education is realized. Many of these activities, in turn, are made up of constituent tasks, that, together and in sequence, work towards the achievement of the goal(s) of the activity. Within this activity structure, discourse functions as one of the means by which the goals of activities and tasks are operationalized. Since an episode is defined as the discourse that mediates the achievement of an activity or of one of its constituent tasks, it is important to provide a characterization of the activity/task from the perspective of the setting it provides for the discourse that occurs. Here, the approach adopted draws on register theory, as developed by Halliday (1978; Halliday and Hasan, 1989) and Martin (1992). Halliday proposes that settings, or situation types, can be described on three dimensions: field, tenor and mode. In the present scheme, field is coded in terms of "episode task", tenor in terms of "participant structure" (3), and mode in terms of "activity orientation". 2. Episode Task Classroom activities and tasks are thought of as falling into the cells of a two-dimensional matrix, where the "subjects" that make up the school timetable form one dimension, and the nature of the students' involvement the other. This latter dimension has four alternatives: Practical This form of involvement is often described as "hands-on"; it involves physical action, often on material objects. The prototypical examples would be science experiments, model making, gym activities, etc. By extension, it is taken to include the activities of writing, drawing and reading, where these are the focus of attention. Discussion This form of involvement occurs when the focus of attention is talk in which there is alternation of speakers. Presentation This form of involvement is more monologic. As well as teacher exposition, it includes activities where one or several students present their work to the rest of the class. Role-Play In this form of involvement, students take on roles appropriate to the topic and

adopt the perspective of the role adopted. In principle, activities corresponding to all the cells in this matrix could occur. However, in this instrument, only those cells that are relevant for our database have been included.

3. Participant Structure In this investigation, only those activities that involve the teacher as one of the participants are being coded. These fall into three subcategories: teacher with whole class, group and dyad.

4. Activity Orientation of Discourse Here, the activity/task is categorized in terms of mode: what role does the talk play in mediating the goals of the activity/task?

C Commenting Talk, descriptive or evaluative, about an ongoing activity.

O Organizing Talk that organizes, directs or monitors an ongoing or about-to-occur activity. P Planning Talk that considers and plans a forthcoming course of action, event or unit of work. T Reporting Talk that narrates or describes a past event (e.g.what we did). Can include report of text read or written. S Problem-Solving Seeking, constructing, testing and evaluating a solution to a problem of a practical, social, symbolic or imaginative kind. G Generating Talk to brainstorm, generate and gather ideas. L Launching Talk to arouse interest and relevant information for a new topic. B Constructing Developing an account, argument or explanation orally. F Formulating Talk to develop an account, argument or explanation in the form of a written, visual or other permanent representation. M Monitoring Talk concerned with checking work completed, the processes used or the skills, information or understanding retained. R Reviewing Talk that retrospectively reviews an activity or unit completed to discover What has or has not been achieved, what issues remain, and/or what processes proved successful or problematic. This is likely to involve adopting a ‘meta’ stance at least part of the time. 7. Episode Development The purpose of this category is to describe the relationship between successive sequences in the development of an episode. The code describes the way in which the current sequence is related to previous sequence(s).

N New Topic (typically at the beginning of an episode). F Further (parallel) contribution on same topic. A Adds to previous sequence through example, anecdote, etc. E Extends previous sequence by developing, justifying or generalizing it. S Consolidates previous sequence(s) in summary, conclusion, etc. C Challenges or counters previous sequence ("yes, but..") R Shifts to related topic, making connection explicit. O Relationship to previous sequence vague and inexplicit.

8. Student Link The purpose of this category is to identify occasions on which a student explicitly makes reference to a previous contribution and to describe the nature of the connection. A Adds to a previous contribution with an example, anecdote, etc. C Challenges or counters it with an objection or alternative. E Extends it by developing, justifying or generalizing it.

O Makes reference but does not make connection explicit.

10. Cognitive Demand Nystrand et al. essentially distinguish between ‘high’ and ‘low’ level demand, with further

subcategories of each. The demand level is judged according to the form and the basis of the answer

that is intended or expected.

Thus, the cognitive level of questions cannot be judged altogether from words alone. For example, if

the teacher expected students to answer questions by reciting information found in textbooks, we

coded questions as reports regardless of their linguistic structure. Hence, though a why-question will

normally elicit an analysis, it will elicit a report (= “rote recall” or “Memory/Prior Knowledge” or

based on “ Previous Conversation”) if the teacher’s focus is the recitation of a textbook's analysis

rather than the class’s reflection or a student’s understanding; then “Why?” really means,

“According to your text, why did it happen this way? Do you remember?” In this instance the

teacher is seeking only recitation. These are all low level demands.

High level demands call for:

Generalizations display inductive reasoning, building up ideas rather than breaking them

down. They address questions such as: What happens? What do I make of what happens? They tie

things together, and they are not restatements of information.

Analyses display deductive reasoning, breaking concepts, ideas, and arguments down rather

than building up ideas. To be scored as analyses, questions had to require more than restatements of

known information.

Speculation considers possibilities, going beyond the information given.

Information about personal experience seems to fall outside this scheme. It should be coded simply as Information. 12. Function of Initiating Move The opening move in triadic dialogue makes a Demand, the response to which can be followed up in various ways. Demands for both Information and Goods/Services seem likely to elicit different types of response depending on whether what is demanded is: Assumed Known, i.e. the response can be evaluated in terms of criteria of conventional correctness, which the student is held responsible for knowing. Personal, i.e. the student is invited to respond from his/her own personal perspective, which is treated as sui generis. Exploratory, i.e. the response is interpreted in terms of its contribution to the ongoing joint activity; while there are no criteria of 'correctness' against which it may be evaluated, the responder may he held accountable for justifying its relevance or appropriateness. The purpose is to negotiate towards concensus. The same distinctions can be applied to initiating moves that are Give in prospectiveness. The following are the subcategories recognized: Information: Assumed Known F Fact: Information of a public kind which the student(s) are expected to know and be able to remember.

A Rule-governed answer: Solution to a problem of an algorithmic kind, as in arithmetic, which the student(s) are expected to know or be able to compute. J Conventional explanation: Culturally recognized explanations in terms of cause, reason, etc., which the student(s) are expected to see as being appropriate. R Report of public event: An account of an event which the student(s) are expected to be familiar with. L Connection: Connections, other than of an explanatory kind, that the student(s) are expected to be able to make. Information: Personal E Experience: Events in the speaker's life that are not assumed to be known to other participants. I Imagination: A personal response of an imaginative kind to a situation or work of art, etc. N Opinion: What the speaker personally believes about a topic or situation of personal concern. Information: For Negotiation Individual contribution to a group attempt to find a solution or reach a conclusion not known in advance. O Opinion: What the speaker personally believes about a topic or situation. P Prediction: What the speaker personally believes or conjectures will happen in an experiment or narrative. X Explanation: Speaker's personal explanation (as opposed to an attempt to repeat the conventional explanation) for an event or state of affairs. C Conjecture: Information that the speaker judges to be relevant to the topic under discussion. L Connection: A link to another topic that is thought to be relevant for the discussion in progress.

Goods and Services: Assumed Known D Act: Perform an action of a conventional kind, which the student(s) can be expected to manage. Goods and Services: Personal G Intention: An action that the student intends or wishes to perform. Goods and Services: For Negotiation S Suggestion: A proposal concerning an action to be performed. Clarification Q Clarification request. At any point in a sequence after the initiating move, a next speaker may make a request for clarification with respect to the preceding move. This has the effect of initiating an embedded exchange which, when the problem is resolved, may be followed by a return to the point in the sequence that had been reached before the clarification request was made.(See also columns 17 &18) 16. Responder In addition to identifying the role of the person providing the response in the exchange, this column is used to note non-canonical ways in which the response slot may be filled. P Pivot: In a nuclear exchange initiated by a Give, instead of explicitly providing the anticipated acknowledge, the next speaker may implicitly acknowledge the previous give by making an initiating move in a new and (usually) related sequence. F Feedback: In the same situation, the next speaker may make a contribution that functions more like one of the subcategories of Follow-up. In this case, an F is entered under responder, and the move described in the columns for follow-up. Q Instead of providing a response, the responder requests clarification. Z In responding to a request for clarification, the responder effectively returns the conversation to the point reached before the clarification request was made.

17. Response

A Acknowledge: the typical, minimum, response to an initiating move that

Gives Information.

Y Confirm/Disconfirm: A polar (yes/no) response, or any synonym of such, that

either confirms/disconfirms or agrees/disagrees with the proposition in the

preceding move.

U Stall: The responder avoids answering the question, typically by “Well..”

Information: Assumed Known

F Fact: Information of a public kind which the student(s) are expected to know

and be able to remember.

A Rule-governed answer: Solution to a problem of an algorithmic kind, as in

arithmetic, which the student(s) are expected to know or be able to compute.

J Conventional explanation: Culturally recognized explanations in terms of

cause, reason, etc., which the student(s) are expected to see as being

appropriate.

R Report of public event: An account of an event with which the student(s) are

expected to be familiar.

L Connection: Connections, other than of an explanatory kind, that the

student(s) are expected to be able to make.

Information: Personal

E Experience: Events in the speaker's life that are not assumed to be known to

other participants.

I Imagination: A personal response of an imaginative kind to a situation or

work of art, etc.

Information: For Negotiation

Individual contribution to a group attempt to find a solution or reach a conclusion not known

in advance.

O Opinion: What the speaker personally believes about a topic or situation.

P Prediction: What the speaker personally believes or conjectures will happen in

an experiment or narrative.

X Explanation: Speaker's personal explanation (as opposed to an attempt to

repeat the conventional explanation) for an event or state of affairs.

C Conjecture: Information that the speaker judges to be relevant to the topic

under discussion.

L Connection: A link to another topic that is thought to be relevant for the

discussion in progress.

Goods and Services: Assumed Known

D Act: Perform an action of a conventional kind, which the student(s) can be

expected to manage.

Goods and Services: Personal

G Intention: An action that the student intends or wishes to perform.

Goods and Services: For Negotiation

S Suggestion: A proposal concerning an action to be performed.

Clarification

Q Clarification : Instead of providing a response, the responder makes a request for

clarification.

At any point in a sequence after the initiating move, a next speaker may make a

request for clarification with respect to the preceding move. This has the effect of

initiating an embedded exchange which, when the problem is resolved, may be

followed by a return to the point in the sequence that had been reached before the

clarification request was made. Clarification may be requested as:

Repetition: Requests the previous speaker to repeat what s/he has just said, either in

the same or similar words.

Identification Requests the previous speaker to identify unambiguously what s/he

intended to refer to.

Confirm: Requests a confirmation/denial of the truth or validity of something that

was previously said (often realised as a tag or rising intonation on the current

speaker's utterance).

Z In responding to a request for clarification, the responder effectively returns the

conversation to the point reached before the clarification request was made.

19. Prospectiveness of Follow-up Moves As with other moves, Follow-up moves can be described by a combination of Prospectiveness and Commodity, with further subcategorization of Commodity. In this column, only the effective prospectiveness of the Follow-up is coded, i.e. that of the segment that selects the next move in the sequence (if any) or that closes the sequence. Prospectiveness: A Acknowledge: Indicates that the previous utterance has been taken into account by the speaker. This is a minimal F move and is quite rare. It might be realized by "OK" or "Yes" said without evaluative overtones. G Give: The F move provides a response to the previous responding move. It is thus, technically, the Initiation of a Dependent Exchange. However, it rarely expects or receives a responding Acknowledge. D The F move, in responding to the previous response, calls for a further response from the same or a different speaker. Depending on the Commodity requested - either a repetition or clarification of the previous response, or some further contribution, the Demand initiates either an Embedded or a Dependent Exchange. G+ A Follow-up move that starts like a comment may have a tag appended or be given final rising tone, which effectively makes the move a Demand. Like a Demand, a Give Plus expects a response. 20. Evaluation Level Nystrand et al. distinguished ‘high’ and ‘low’ level evaluation, with ‘high’ involving some form of

‘uptake.’

Uptake. The issue here concerns the way in which the previous response is taken up. Is the response

given recognition in its own right by being included in some way in the follow up move? Uptake

can be realized in a comment that explicitly recognizes the previous response and builds on it in

some way; or it can be realized in a question that asks the previous responder to extend what s/he

said.

[We defined uptake as occurring when one conversant, e.g., a teacher, asks someone else, e.g., a

student, about something the other person said previously (Collins, 1982).

Uptake is often marked by the use of pronouns, e.g., “How did it work?”, “What caused it?”, “What

city grew out of this?” In each of these questions, the italicized pronoun refers to a previous answer.

To qualify as uptake, a question must incorporate a previous answer, not a previous question; hence,

we did not code as uptake teachers making reference to questions or remarks they had previously

made.

Do not consider as high-level a teacher's introduction of new information in response to a student

answer unless the teacher incorporates a previous student answer; the criterion is the importance of

the student as a source of new information.

We coded teachers’ evaluation of student responses as high when the student contributed

something new (i.e., new information) that changed or modified the topic of discourse in some way,

and was acknowledged as such by the teacher. In other words, when a teacher's evaluation is high-

level, the student really “gets the floor.” Specifically, we operationalized high-level evaluation using

two criteria: (a) the teacher's certification of the response (“Good,” “Interesting,” etc.) and (b) the

teacher's incorporation of the response usually in the form of either an elaboration (or commentary,

e.g., “That’s important because . . .”) or a followup question (e.g., “Can you say more about that?” or

“Why do you say that?”). That is, for level of evaluation to be coded as high, the evaluation had to

be more than “Good,” “Good idea,” or a mere repeat of a student's answer. In all instances of high-

level evaluation, the teacher validated the student's answer so that it affected the subsequent course

of the discussion. ]

Accept + Uptake: as described above

Accept + Praise: More than mere acceptance; the praise is explicit.

Reject + Justification: As well as rejecting the preceding response, the speaker gives a reason.

Accept/Reject: Indicates that that the information provided or action performed is thought to be

appropriate or inappropriate by the speaker.

21. Function of Follow-Up: Commodity: F moves that are G or D in prospectiveness introduce new material into the exchange in which they occur. The possibilities recognized are: V Evaluation: Expresses or calls for an opinion about how useful or appropriate another person's contribution (in words or action) has been. J Justification/Explanation: Provides or requests an argument to support a preceding contribution or to show that it was reasonable or appropriate. C Comment: Makes or requests a comment that extends or modifies a preceding contribution. U Clarification: Requests a repair with respect to the previous contribution. A Action: Makes a request referring to action.

M Metatalk: Treats the discourse as the topic for discussion. Evaluation: The subcategories of Evaluation are: A Accept: Indicates that that the information provided or action performed is thought to be appropriate by the speaker. R Reject: Indicates that the information/action is thought to be inappropriate by the speaker. C Correction: Indicates that something previously said or done is considered to be wrong, and expresses the speaker's view of how the error should be rectified. F Reformulate: Expresses in different words (often more concisely or completely) something that was said by a previous speaker. D Counter: Expresses disagreement by offering another argument or an alternative interpretation of a point made by a previous speaker. Y/N Repeat: The speaker reiterates what someone has previously said, either in the same or very similar words in order either to accept or reject the move in question, with the evaluation signalled by intonation and by what is done in the following move. Code Y for Accept or N for Reject. Comment: The subcategories of comment that may be given or requested are: E Exemplification: Provides or requests an example to illustrate a point that has just been made. A Amplification: Provides or requests a filling out or qualification or modification of something previously said in terms of temporal, spatial, causal, or conditional detail (Eggins and Slade, 1997, p. 198). C Connection: Provides or requests additional information to supplement what has just been said, or to connect it to some other domain, with the effect of developing the topic of the current sequence. S Summarize: Provides or requests a condensed statement of what has been said by one or more participants. This is often done in such a way that the

other participants understand that closure has been provided and the topic should not be further pursued. The following only occur in the Demand mode: O Opinion: Requests another participant(s) to express his/her personal views or conjectures on the topic under discussion. D Description: Requests another participant(s) to tell or write about a physical object or abstract idea. L Observation: Requests another participant(s) to describe something that s/he is looking at or has recently seen. Action: The subcategories of Action are: N Requests action immediately. F Requests action in the reasonably near future. S Suggests action in the future. Clarification: The subcategories of Clarification that may be requested are: R Repetition: Requests the previous speaker to repeat what s/he has just said, either in the same or similar words. I Identification Requests the previous speaker to identify unambiguously what s/he intended to refer to. C Confirm: Requests a confirmation/denial of the truth or validity of something that was previously said (often realised as a tag or rising intonation on the current speaker's utterance). Metatalk: C Metacognitive: Refers to the procedures and strategies used in the discourse. O Metaorganizational: Refers to the organization of the activity.

T Metatopic: Refers to characteristics of the under discussion as a whole. 28. Sequel This column is used to describe the type of exchange that immediately follows the follow-up. Apart from continuations of the exchange initiated by a follow-up with Demand prospectiveness, the categories here are the same as in column 7.

References Bakhtin, M.M. (1986) Speech genres and other late essays. Austin: University of Texas Press. Edwards, D. and Mercer, N. (1987) Common knowledge: The development of understanding in the classroom. London: Routledge. Eggins, S. and Slade, D. (1997) Analysing casual conversation. London: Cassell. Engeström, Y. (1991) Non scolae sed vitae discimus: Toward overcoming the encapsulation of school learning. Learning and Instruction, 1: 243-259. Halliday, M.A.K. (1978) Language as social semiotic: The social interpretation of language and meaning. London: Arnold. Halliday, M.A.K. and Hasan, R. (1989) Language, context and text: aspects of language in a social semiotic perspective. Oxford: Oxford University Press. Halliday, M.A.K. (1993) Towards a language-based theory of learning. Linguistics and Education, 5: 93-116. Kamberelis, G. (1995) Genre as institutionally informed social practice. Journal of Contemporary Legal Issues, 6: 115-171. Lemke, J.L. (1994) Intertextuality and educational research. Linguistics and Education, 4 (3-4): 257-268. Leont'ev, A.N. (1981) The problem of activity in psychology. In Wertsch, J.V. (Ed.) The concept of activity in Soviet Psychology. Armonk, NY: Sharpe. Martin, J.R. (1992) English text: System and structure. Amsterdam: John Benjamins. Matusov, E. (1996) Intersubjectivity without agreement. Mind, Culture, and Activity, 3 (1): 25-45.

Nystrand, M., Wu, L.L., Gamoran, A., Zeiser, S. and Long, D. (2000) Questions In time: Investigating the structure and dynamics of unfolding classroom discourse. National Research Center on English Learning and Achievement (CELA).The University of Wisconsin-Madison. Wells, G. (1993) Reevaluating the IRF sequence: A proposal for the articulation of theories of activity and discourse for the analysis of teaching and learning in the classroom. Linguistics and Education, 5: 1-37. Wells, G. (1996). Using the tool-kit of discourse in the activity of learning and teaching. Mind, Culture, and Activity, 3 (2): 74-101. Wertsch, J.V. (1994) The primacy of mediated action in sociocultural studies. Mind, Culture, and

Activity, 1 (4): 202-207. Gordon Wells 22 October 1997. (Revised November 2001.)